M.J. Welsh and A.E. Smith, RAC Application 
conductance and the capacity for vectoral transepithelial electrolyte transport). These are 
the characteristics which define an epithelium. 
An example of their value as a model system for assessing therapy is that this system has 
led to the discovery and subsequent in vivo clinical testing of two potential therapies for 
CF: amiloride (17) and extracellular nucleotides (19). In addition, the ability to measure 
the transepithelial conductance provides the most sensitive assay of cell and monolayer 
integrity; with injury of epithelial cells, transepithelial conductance is an early, perhaps the 
First, parameter to increase (135,136). 
We believe that human airway epithelia grown on permeable supports provide the model 
which most closely resembles human airway epithelium in vivo. 
3. Nasal and bronchial epithelium of Rhesus monkeys. 
This model system has the advantage of being an in vivo model, and it resembles the 
proposed human experiment. It has the disadvantage that the monkey has wild-type CFTR 
function; expression of recombinant CFTR on a wild-type background produces little, if 
any, functional evidence of CFTR expression. We have documented this by expressing 
CFTR in normal human airway epithelial cells; we found no additional increase in chloride 
secretion. Similar results have been reported by others (66). The lack of an increase in 
chloride secretion with overexpression of CFTR on a wild-type background likely results 
from limitations in transepithelial secretion resulting from rate-limiting steps occurring at 
the basolateral membrane (see Section 3.4). Alternatively, there could be rate-limiting 
steps in terms of insertion of CFTR into the apical membrane or in the signal transduction 
mechanisms that regulate CFTR chloride channels. As a result of these considerations, we 
have also used the Ad2/6Gal-l vector to assess the efficacy and safety of the delivery 
system. We also applied Ad2/BGal-l to the bronchial epithelium of the monkeys to 
demonstrate the ability of the virus to express recombinant protein in lower airway 
epithelium. 
4. CF Mice 
Two groups of investigators have recently reported the development of CF mice by targeted 
disruption of the endogenous mouse CFTR gene (137-139). We have not yet tested 
Ad2/CFTR-1 in such mice. However, we would argue that such studies are not essential 
prior to testing Ad2/CFTR-1 in humans, a) The use of CF mice is unlikely to provide new 
data relevant to safety, b) The current models do not yet show lung pathology and thus do 
not yet provide a model of human lung disease, c) The outcome, either positive or 
negative, of studies directed at determining whether Ad2/CFTR-1 can express protein and 
restore cAMP-regulated chloride in mouse airway epithelium, would not influence our 
conclusions about the potential efficacy of the vector, based on our studies of efficacy in 
human CF airway epithelia and in primate nasal and bronchial epithelia. 
B.2.b(2) What is the minimal level of gene transfer and/or expression that is estimated to 
he necessary for the gene transfer protocol to be successful in humans? How was this 
level determined? 
As we discussed above in Section 3.3, we estimate that very little CFTR need be expressed 
to successfully correct the chloride transport defect in CF airway epithelia. This conclusion 
is based on both theoretical considerations and experimental observations. 
[ 888 ] 
Recombinant DNA Research, Volume 16 
